14 replies to this topic

[bobmeetin]

Please point me to a relevant thread and close this cookie if another thread answers this question:

I'm evaluating a project in which the business has a main business site, then links to 4 sub-domains for each product area they offer. The main site has marginal content, about us, a sort-of testimonial page, a newsletter index, a badly broken sitemp (Warning: include() [function.include] ...... you get the picture), and of course links to the 4 sub-domains, product sites.

Navigation-wise it's a disaster, a nightmare borne from a nightmare. The product sites have similar structure (different color themes) amongst themselves, but the main site is little more than a landing page.

I can see that a lot of work was put into setting up the separate product sites, albeit I don't think they are particularly user-friendly. There is substantial content, almost too much in my opinion. The left navigation goes on forever with links.

I won't go into details (no site name) but one of the product pages has to do with pre-diabetes. You have to go several pages deep into google to find any natural results. They do some adwords - and that is pretty inconsistent.

Pagerank, if it can be trusted, is commonly 2 or perhaps 3.

So what am I asking? The seemingly low ROI based upon search engine findability, how much can be attributed to the use of sub-domains vs simply poor organization?

I'm not keen on the sub-domains, but I could certainly do a restructuring of the home page to give it similar look/feel as the sub-domains so that moving around is not so awkward. And fix what is broken.

-Bob

[torka]

So what am I asking? The seemingly low ROI based upon search engine findability, how much can be attributed to the use of sub-domains vs simply poor organization?

Almost entirely to poor organization, IMO.

I manage a website every day that's made up of two different domains plus a subdomain. Each domain/subdomain comprises a different "section" of the corporate site: the main/informational section, the customer support section and the web store. Pages from the domains/subdomain all do well in the SEs.

What we do is treat the pages as though they all belong to a single site (because they do). For instance, the "Home" link always links to the default page for the main domain, no matter which domain the page you might be looking at resides on. The "Products" link always points to the default page of the store domain, no matter what domain the page you might be looking at resides on. The "Support" link always points to the default page of the support subdomain, no matter what domain the page you might be looking at resides on.

In other words, the top-level site navigation integrates the domains/subdomain as a single site made up of multiple sections. Some of those sections are folders within the "main" domain; one is another domain; another is a subdomain. From the standpoint of our navigation, it makes no difference whether it's a folder or another domain -- we treat them the same.

Within each of those sections -- particularly within the store -- there is section-specific navigation. And there are in-line text links from the content of various pages pointing toward important pages throughout the site, again regardless of which domain the linking page and the target page reside on. This helps the SEs figure out which pages are most important on each of these domains/subdomain.

IMO, what you really have there is a single site made up of a domain and four subdomains. So it's just a matter of setting up the sitewide navigation accordingly. It really doesn't matter if the subdomains look the same, or if they use the same basic template but different colors, or if they're entirely different looking.

It's your navigation that will determine whether what you've got is one big, strong site or a collection of smaller, weaker "mini-sites."

--Torka

[bobmeetin]

I'm with you on all this. When I look at the main site, then the sub-domains, it is disorienting. Technically I'm pretty adept with standardization, navigation, but what is a little unclear in this case, is how much better off they would be if they/we fixed it up. It's tough to quantify.

The sub-domains sites are standard 3 column structure. They have some quotes, testimonials in the right column, I think fairly done, but the left column just goes on forever. It repeats page by page which sort of seems fine, but really is a bit tough to visually follow.

Left column example:

About us:
link 1 - imagine this link is a sentence that breaks to 3 lines!
link 2
link 3
...
...
Resources:
link 1
link 2
link 3
...

3rd category:
link 1
link 2
link 3
...
4th category - etc......
5th category - etc...

So the left column has lots of categories and lots of very long sentence links. It feels to have been built for search engines, not people visitors. In the site design they a chose "soft" effect. i.e. a soft background color behind text and the text is lighter than black.

I find myself questioning the design choice itself and getting distracted as to whether I am simply picking needlessly in the wrong direction. IMHO - way way too many long links to be people friendly. -Bob

[torka]

Sounds pretty dire.

They might want to consider some kind of collapsing menu so all those links wouldn't be displayed at once. They might also want to consider shortening the menu links. In my experience, it really isn't necessary to have sentence-long links on every page of the sub-section in order to get the target page ranked for those kind of longer phrases. Usually those phrases are less competitive, and an inline text anchor link or two on a couple of category level pages might be enough.

If it doesn't seem people-friendly to you on first glance, it probably isn't. A little usability test may be in order. Bring in a couple of "ordinary people" (friends, family members if available) to take a look at it. Give them a couple of tasks to do that would require them selecting links out of the list on the left (but don't tell them the exact text of the link or what link group/section they'll find them in), have them talk out loud what they're thinking while they're doing it (they'll feel a little self-conscious about this, but it's really useful for you to hear what's going on in their head) and see how easy or hard they find it to complete those tasks. If they go directly to the appropriate link(s) without any hesitation or trouble, then you can feel somewhat more comfortable that you might be overreacting... but if they have to hunt around or have troubles finding link(s) they want, you'll know you're on the right track.

--Torka

[bobmeetin]

I asked one of my friends to check out the site (which you know I can't identify here) and she agreed with regard to the site being 'complex'. Although it wasn't communicated directly to me by the business owner I heard that they were planning to have the site redeveloped. And hopefully - making it people-friendly will be at the heart of their discussions.

There are myriads of navigational options that could make it more pleasing to the audience. Since much of the stuff hanging out in the left column is really "How-to" links I could see moving the whole batch into a main content Q/A page and perhaps either adding some sort or search options to the page itself.

And also perhaps getting rid of some of the walking billboard stuff.

-Bob

[torka]

Sounds like you're on the right track with your thinking.

Perhaps communicate to the business owner how important well-organized navigation is to the success of an optimization campaign, how much easier it is to optimize when proper optimization is "baked in" from the start -- try to get yourself a seat at the table when they're planning/developing the revamped website.

Good luck!

--Torka

[Randy]

I fully agree with Torka.

Since they're already considering a redesign project now would be the perfect time to make sure your two cents become part of the institutional thought process. Who knows, if someone is waffling about whether to take on a redesign/cleanup effort your gentle nudging may even be the emphasis needed to get the project started!

One thing, since it's something I forget myself all too often when I put on my design hat and take off my marketing hat...

Make sure you know what the USP/UVP (Unique Selling Proposition/Unique Value Proposition) is before the redesign starts, then make sure the final design supports this USP/UVP in a clear way.

As an example of how this can go all wrong, I have a site whose USP is basically that the service is provides it's very easy to use. And it is. eg save money by saving time. It's a good USP. But I made a mistake in redesigning the site almost a year ago in that the final design made it appear to be a lot more complicated than it is. I'd included too much information in the pre-buy, public area pages rather than keeping it as simple as it really is..

Bad, bad Randy. I know better!

I'd been wondering why sales and conversions weren't continually rising as they have been for years now as the offer improved and the site gained more authority. But it just didn't make it onto my radar screen because the site's conversions overall were still pretty decent and revenue was slightly up from historical levels.

The lack of a bigger jump finally nagged at me enough that I looked at it again a month or so ago, had some others look at the site. Together we spotted the juxtaposition of what I was saying in the copy and what the design was saying. I then came up with a quick split test with some radical redesigns that affect everything from colors to navigation options to text on the landing pages. I'm not nearly done testing it yet, but already I'm seeing a 35% or so improvement in conversions with a couple of my simplified versions.

Lesson learned.

[bobmeetin]

It's funny but one of the problems with being technically inclined, not even at guru status, is that sometimes you try to make a solution too perfect, and along the way, you put in more detail than can be handled or is necessary. That happens here as well, so the art is really in recognizing the signs. In Luv-n-Logic they call it over-splaning.

Early this AM I emailed the business owner type a low-level evaluation of the site design, navigation, usability, etc. with minimal detail talking to this thread summary. He was either not shocked or exercising diplomacy but the bottom line is that he needs to be careful of overly alerting the current designer, etc. before the time is ripe.

This all adds up to "What a mess!" Now I sit back with the negotiator hat at arm's reach.

[Randy]

Sounds like about the best you can do at this point Bob. But at least it appears as if you might have a seat at the table when the redesign talks start.

On the other stuff, I really do know better. Heck, it's like step #3 when I'm starting a new site. (1 is figuring out who the target audience is, 2 is how what I offer can benefit them, which leads to 3 being defining a strong USP and working everything towards supporting that.) This was an older site though, so I already knew 1 and 2. For some stupid reason I completely skipped step 3. And was apparently too close to the redesign project to see what I'd actually done. Forest, trees and all of that.

That's okay though. I'm sure it'll lead to a good article or guide when I find a little free time. We do learn from our mistakes. Especially the ones that cost us a significant amount of money in lost sales.

[Firstwebsearch]

I haven't looked at the site so this is based off my understanding of your post only. Though subdomains can be useful in SEO in certain situations, they are only useful if the sites architecture is basically flat. It sounds as if once you get into a subdomain there are dozens of levels of navigation under that which would kill seo efforts. You should get them to redesign the site in as flat an architecture as possible, increase the landing page (main domain) as much as possible with relevant content and use the subdomains with a max of two clicks to any area. As for navigation I would agree that a collapsible menu would be ideal.
If done right you could end up with control of up to 4 spots on any SERP. That is because the subdomains are looked at as different sites.

[Randy]

If done right you could end up with control of up to 4 spots on any SERP. That is because the subdomains are looked at as different sites.

Not anymore. At least not anymore and not for awhile where Google is concerned, except for in some very special circumstances.

As of last December they started treating subdomains like they do subdirectories where host crowding is concerned.

[bobmeetin]

Firstwebsearch - I'm not clear on what you're describing as flat. I misplaced my tech hat I think. If you're are talking to resynchronizing the site+4 from top to bottom, look/feel/navigation then I'm with ya. What I am currently seeing (and it looks like my good friend, PHP) is the main site with a few almost pointless about type pages, and 4 links going to the subdomains. Without going too deeply into unpaid consulting, I did a quickie link check through marketleap and find really inconsistent results:

main: 800+ total
subA: 52 total
subB: 5500 total
subC: 70 total
subD: 300+ total

Unless I am misunderstanding, this says to me that they're performing as independent silos, that there is a lot of unrealized potential.

[bobmeetin]

To any, to all - the link checking tells me that subdomains A and C are a little shaky. Reorganizing to a single domain would certainly (yes) unify the link back situation.

Q1: Aside from the time/effort involved and if it was not a huge issue, is there a reason not to do this? Technical or search-engine wise? Part of the process would have to include setting up permanent redirects.

Q2: Other than doing a link exchange campaign is there another option (creative is fine) to solve the link-light problem that does not involve reinventing the wheel?

Cheer-i-o

Edited by bobmeetin, 28 October 2008 - 10:07 AM.

[Randy]

I wouldn't personally be paying much attention to the raw count of links Bob. And if I did I'd make sure I was comparing apples to apples so that all of those were only looking at External links. With External meaning I'd be removing any links pointing from either the main domain or any of the subs.

Trying to put too much stock in link counts in a situation like this could lead you down a false path. Not necessarily a bad path, just one where you don't get the real info you need. Especially since raw link counts isn't nearly as important as the quality of those links.

Long story short, I'd first work on cleaning up the navigation. Setting things up where the subs are treated basically as parts of the main site, so that you really end up with a single site that has different divisions. I highly suspect once that's done a lot of the problems will magically go away, because you'll be using your internal linking structure to pass the link popularity properly.

I'd do this first because I have a sneaking suspicion that this clean up will have a positive effect across the board. Basically because you'll be taking the link popularity of the main and the four subs and combining them or merging it all into a single site structure.

That's what it sounds like you actually have. But because of the wonky navigation it's not being made clear to the search engines that it's all connected. Working out a new navigation structure should help both the search engines and real users figure out what's really happening.

[torka]

Wot Randy said.

Work out the site architecture/internal linking structure and let things cook for awhile before you make any other significant structural changes. I, too, suspect once you work out the internal navigation to let the SEs know not only that all these things are related, but also how they're related, you'll see a lot of the current issues clear up.

Don't worry so much about which page resides on which specific subdomain. Just mentally translate those subdomains into folders and link pages in them back and forth as appropriate, as you would if they were all on one domain. Fix up/standardize the visual design as needed to keep from freaking out the human visitors and you'll be good to go. Honestly, if the visual design is coordinated, the vast majority of the human visitors will never notice if/when they've gone from one subdomain to another as they navigate around.

What you should have when you're done is a large site with a lot of inbound links pointing not only to the home page but also to many interior pages -- which is a good thing -- that just happens to include pages hosted on one main domain and four subdomains. Perfectly acceptable practice, and one that I can tell you from experience will work just fine for all the major SEs.

Improvement won't happen overnight -- nature of the beast with SEO -- so the client needs to be prepared to exert a little patience. But I really think starting with sorting out the site architecture and internal linking structure is the way to go.

My

--Torka